Unravelling Power of the Unseen: Towards an Interdisciplinary Synthesis of Generative AI Regulation

Abstract The regulations of generative AI, typified by ChatGPT and Sora, have become one of the most influential alternative technological imaginaries. Developed by states and civil society groups, such regulations are triggering a broad range of social actors seeking to nominalize the AI-related behavior. Against this backdrop, this study starts with interrogating the semiotic character of generative AI. Do these regulations support the AI futures, or do they involve a mere change in the social actors who benefit from the technological status quo? To answer this question, this study examines the rhetoric and realization of AI regulations by the European Union and the United States. The findings reveal a degree of AI regulatory alignment between the European Union and the United States, but these two jurisdictions also highlight and predict some structural challenges. Drawing upon the concept of panopticism by Foucault, the study explores the foundational origins of challenges by dissecting the (in)visibility of AI power. It underscores the necessity of regulating the power of the unseen and proposes a synthetic generative AI regulatory framework. We finally conclude that the integrity of sociosemiotics and panopticism provides a productive and paramount framework for understanding the powerful new capacities of AI-related regulations.